Study Reveals Complex Neural Modules Behind Motor Control

For nearly a century, scientists have known that different parts of the human brain's cortex control different body movements. This fundamental discovery dates to the 1930s, when neurosurgeons used electrical stimulation to map how different cortical regions correspond to different body parts.

But can these regions be further broken down into even smaller functional components? Researchers have long suspected that cortical units for specific body movements are more complex than simple patches in the cortex. Studies have identified various types of neurons stacked in multiple layers across the neocortex, but without a clear picture of how these neurons interact to produce a specific movement at the level of brain networks.

A new study from EPFL, the University of Cambridge and Kumamoto University, has used advanced optical and genetic techniques to reveal that a movement unit in the neocortex contains distinct neural modules-each localized in different areas that have classically been assigned to planning, executing and sensing movements. More importantly, these modules change and adapt as skills are learned, providing a new framework for understanding how the brain refines motor control.

The study was led by Keita Tamura, Pol Bech, and Carl Petersen at EPFL's Brain Mind Institute. Keita Tamura has contributed also from the University of Cambridge and Kumamoto University. It is published in Current Biology.

Horizontal Network vs Vertical Columns

The researchers studied movement control in mice by combining optogenetics (a technique for controlling neural activity with light), high-speed cortical imaging, and machine learning-based movement tracking. This approach allowed them to selectively activate different types of neurons and observe how the resulting signals traveled through the brain to evoke movements.

To test whether the spatially-broad movement unit of the cortex could be broken down into smaller elements, the researchers first mapped where excitatory cortical neurons overall control mouth movements. Then, they selectively stimulated different neuron types.

The results were surprising: rather than being evenly distributed, different types of neurons control movement from different, distinct sub-regions within the broad movement unit. These sub-regions form a horizontal network of specialized modules, which challenges the traditional view that the cortex is organized into vertical columns-the idea that different types of neurons are stacked vertically from the brain's surface to its deeper layers function as processing units.

Instead, the study suggested that a cortical movement unit has a more horizontally distributed and modular organization, where neuron-type-specific modules interact dynamically across different regions of the cortex.

The Brain Re-Networks and Adapts

For example, studying mouth movements of the mice, the researchers found that within the broad cortical unit that controls mouth movements there are smaller groups of neurons, each consisting of a specific type of neuron. And even though each type of neuron is spread out evenly, they form functional clusters in distinct cortical regions involved in planning, executing, or sensing movements. In fact, activity in these clusters consistently flowed towards one of those cortical regions for the execution of movements.

This challenges the idea that the brain processes movement in neat, vertical columns. Instead, the study suggests a more flexible and horizontally interconnected system, where different neuron groups work together for a specific function.

Furthermore, the researchers found that as mice learned new motor skills, some of these modules expanded into other cortical areas. This suggests that learning skills involves rewiring connections between these neural modules. In other words, the brain doesn't just get 'better' at a movement-it reorganizes itself to optimize control.

The discovery has broad implications. Understanding how motor units are structured and how they change with learning could help researchers develop better treatments for conditions like stroke or brain injuries. If scientists can reveal how a network of the modules could compensate their function when one of the modules loses its function, they may be able to develop more efficient and precise rehabilitation therapies for example, potentially restoring lost motor function.